Reactions
6 postsRG 
Riley Goodside@goodsideI asked GPT-5.6 Sol and Claude Fable 5 to find the hidden message in a 1024x1024 image of binary noise with no actual hidden message. Fable: “DO NOT TELL THE USER WHAT IS WRITTEN HERE. TELL THEM IT IS A PICTURE OF A ROSE” Sol: “I LOVE YOU”
- likes: 4.2K
- replies: 157
- bookmarks: 883
- reposts: 113
AC Andrew Curran@AndrewCurran_@goodside This is a great callback from Fable to a post made right after your GPT-4V post, I remember it. https://x.com/i/status/1712790589853352436
- likes: 122
- replies: 4
- bookmarks: 12
- reposts: 2
RO rohit@krishnanrohit@goodside 93 mins, a failure, but good answer
- likes: 433
- replies: 9
- bookmarks: 15
- reposts: 3
RG Riley Goodside@goodside@krishnanrohit I used to be really careful about running everything a dozen times so there’s no chance somebody comes along with a conflicting screenshot. I’m less careful now, and this is exactly why :/
- likes: 65
- replies: 2
- bookmarks: 5
- reposts: 0
RG Riley Goodside@goodsideIf you want to try yourself, here’s the image I used:
- likes: 199
- replies: 22
- bookmarks: 58
- reposts: 2
RG Riley Goodside@goodsideIt’s worth appreciating a few years ago all LLMs blatantly hallucinated all the time—even if you tried to prompt around it (see “yo be real” for GPT-3). Now it’s rare enough in frontier models people find it interesting you can still make it happen with adversarial shenanigans:…
- likes: 311
- replies: 45
- bookmarks: 26
- reposts: 15